🌟 Aurora - AI Ecosystem for Professional Development

An AI ecosystem that empowers young professionals from onboarding to leadership by providing personalized guidance, skills growth planning, and transparent feedback.

Note: Hosted Servers are of Free Tier and may be in sleep mode

🎯 Vision & Mission

Aurora is an AI ecosystem designed to empower young professionals throughout their career journey - from initial onboarding to leadership roles. Our mission aligns with SAP's motto: "help the world run better and improve people's lives" while showcasing Responsible AI and business scalability.

The platform provides personalized guidance, skills growth planning, and transparent feedback to help young professionals navigate their career development.

🏗️ System Architecture

Frontend (Next.js 15)

• React 18 with TypeScript for type safety
• Real-time streaming responses with progressive loading
• Responsive design with Tailwind CSS
• Multi-agent chat interfaces with agent-specific UI
• Dynamic height containers to prevent content truncation
• Loading indicators with agent-specific messaging

Backend (FastAPI + Python)

• Async API with Python 3.9+
• Modular agent architecture with specialized AI agents
• RESTful endpoints with streaming support
• Comprehensive audit logging system
• Health monitoring with detailed system status

AI & Data Stack

• OpenAI GPT-4 for intelligent learning plan generation
• Qdrant for high-performance vector storage and semantic search
• LangChain for RAG pipeline and agent orchestration
• HuggingFace Embeddings (sentence-transformers/all-MiniLM-L6-v2)
• MySQL for structured data storage
• CSV files for resource and progress data

🤖 AI Agents

1. Welcome Agent 👋

Purpose: Onboarding FAQs and policy guidance for young professionals

• RAG-powered responses using company documentation and handbooks
• Policy document filtering for accurate onboarding information
• Source attribution with document references for transparency
• Streaming responses for real-time interaction

Key Capabilities:

• Answers onboarding FAQs from policy documents and handbooks
• Provides guidance on company policies, benefits, and procedures
• Offers transparent, source-backed responses for new employees
• Supports young professionals in their initial company integration

2. Skill Navigator Agent 🧭

Purpose: Personalized 30-60 day learning plans for professional development

• Skills gap analysis with targeted learning recommendations
• SAP Learning Hub integration for relevant course content
• Career-focused planning aligned with professional growth goals
• Adaptive learning paths that evolve with individual progress

Key Features:

• Analyzes user learning requests using natural language processing
• Retrieves relevant company project documentation using RAG
• Generates personalized 4-week learning plans with specific goals
• Recommends resources from company learning catalog
• Connects learning objectives to actual company project requirements

3. Progress Companion 📊

Purpose: Learning progress tracking and motivational feedback

• Progress analysis with intelligent insights and recommendations
• Achievement tracking with milestone recognition
• Motivational feedback to maintain engagement
• Next steps guidance for continued professional development

🚀 Key Features

Personalized Professional Development

• Career-focused guidance from onboarding to leadership
• Skills gap analysis with targeted learning recommendations
• SAP Learning Hub integration for relevant course content
• Transparent feedback on progress and achievements
• Responsible AI practices with explainable decisions

Intelligent Learning Planning

• 30-60 day structured plans tailored to individual needs
• Skills-based recommendations aligned with career goals
• Learning path optimization based on current competencies
• Progress tracking with motivational feedback
• Adaptive planning that evolves with professional growth

Advanced AI Technology

• Qdrant vector database for high-performance semantic search
• RAG-powered responses using company documentation
• Real-time streaming for interactive user experience
• Multi-agent architecture for specialized assistance
• Comprehensive monitoring with audit trails

Comprehensive Monitoring

• Health check endpoints with detailed system status
• Vector store monitoring with document counts
• Agent performance tracking with latency metrics
• Debug logging for troubleshooting
• Audit trail for all interactions

📋 Getting Started

Prerequisites

• Node.js 18+ and npm
• Python 3.9+ with pip
• MySQL database (local or cloud)
• OpenAI API key
• Qdrant instance (local or cloud)

Installation

1. Clone Repository

git clone https://github.com/your-org/aurora.git
cd aurora

2. Backend Setup

cd backend
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate
pip install -r requirements.txt

# Configure environment variables
cp .env.example .env
# Edit .env with your configuration:
# OPENAI_API_KEY=your_openai_api_key
# QDRANT_URL=http://localhost:6333
# QDRANT_API_KEY=your_qdrant_api_key
# QDRANT_COLLECTION=aurora_docs
# AUTO_INGEST=1
# SEED_DATA_DIR=./data

# Start backend server
uvicorn app:app --host 0.0.0.0 --port 8000 --reload

3. Frontend Setup

cd frontend
npm install

# Configure environment variables
echo "FASTAPI_URL=http://localhost:8000" > .env.local

# Start frontend server
npm run dev

4. Access Application

Open browser to http://localhost:3000

📊 Project Structure

Aurora/
├── backend/                    # FastAPI backend
│   ├── agents/                # AI learning agents
│   │   ├── onboarding/        # Welcome agent (RAG-powered)
│   │   ├── skillnav/          # Skill Navigator (AI-powered)
│   │   └── progress/          # Progress Companion (CSV reader)
│   ├── app.py                 # Main FastAPI application
│   ├── rag.py                 # RAG system implementation
│   ├── vectorstore.py         # Qdrant vector store management
│   ├── orchestrator.py        # Agent orchestration
│   ├── settings.py            # Configuration management
│   └── requirements.txt       # Python dependencies
├── frontend/                   # Next.js frontend
│   ├── src/app/               # Next.js app router
│   │   ├── agents/            # Agent interfaces
│   │   │   ├── welcome/       # Welcome agent page
│   │   │   ├── skillnav/      # Skill Navigator page
│   │   │   └── progress/      # Progress Companion page
│   │   └── api/               # API proxy routes
│   │       └── aurora/        # Backend proxy
│   ├── components/            # UI components
│   │   └── ChatStream.tsx     # Streaming chat component
│   └── lib/                   # Utility libraries
├── data/                      # Training data and documents
│   ├── projects/              # Company project documentation
│   ├── policies/              # Company policies
│   ├── resources/             # Learning resources catalog
│   └── seed/                  # Seed data for auto-ingestion
└── README.md                  # This documentation

🛠️ Technical Details

Dependencies

• Core Framework: FastAPI, Next.js 15, React 18, TypeScript
• AI/ML: OpenAI API, LangChain, Qdrant, HuggingFace Embeddings
• Database: MySQL, SQLAlchemy, Qdrant Client
• Utilities: Python-dotenv, Tenacity, Watchfiles, Uvicorn

API Endpoints

• GET /healthz - Comprehensive health check with system status
• POST /agents/onboarding/stream - Welcome agent with RAG responses
• POST /agents/skillnav/stream - AI-powered learning plan generation
• POST /agents/progress/stream - Progress tracking and analysis
• GET /admin/audit/* - Audit trail and monitoring endpoints

Environment Variables

Backend Configuration

# Required
OPENAI_API_KEY=your_openai_api_key
QDRANT_URL=http://localhost:6333
QDRANT_API_KEY=your_qdrant_api_key
QDRANT_COLLECTION=aurora_docs

# Optional
AUTO_INGEST=1                    # Auto-ingest documents on startup
SEED_DATA_DIR=./data             # Directory for seed documents
EMBED_MODEL=sentence-transformers/all-MiniLM-L6-v2
AURORA_DB_URL=mysql+mysqlconnector://user:pass@localhost:3306/aurora_db
AURORA_HMAC_KEY=your_secret_key

Frontend Configuration

# Required
FASTAPI_URL=http://localhost:8000

📈 Usage Examples

Welcome Agent Queries

• "What is the company leave policy?"
• "How do I request vacation time?"
• "What are the company benefits?"
• "Tell me about the onboarding process"
• "What should I know as a new employee?"

Skill Navigator Queries

• "Create a 30-day learning plan for a junior developer role"
• "I need to improve my Python and cloud skills for career growth"
• "Plan my development from junior to senior developer"
• "What should I learn for data science career advancement?"
• "Help me become a technical lead in 60 days"

Progress Companion Queries

• "What's my current learning progress?"
• "Which courses are overdue?"
• "What should I focus on next for career growth?"

🔧 Configuration & Deployment

Qdrant Setup

1. Local Qdrant: docker run -p 6333:6333 qdrant/qdrant
2. Cloud Qdrant: Use Qdrant Cloud with API key
3. Collection: Automatically created with dimension inference

Data Setup

1. Place company project documentation in data/projects/
2. Update learning resources in data/resources/catalog.csv
3. Configure user progress data in data/courses.csv
4. Set AUTO_INGEST=1 for automatic document processing

Production Deployment

• Use environment variables for all configuration
• Enable SSL for database connections
• Set up proper logging and monitoring
• Configure backup strategies for vector store

🔍 Recent Improvements

Professional Development Focus

• ✅ Career-Centric Design: Aligned with SAP's mission and values
• ✅ Young Professional Focus: Tailored for onboarding to leadership journey
• ✅ Responsible AI: Transparent, explainable decision-making
• ✅ Business Scalability: Architecture designed for enterprise growth

Technical Enhancements

• ✅ Qdrant Integration: High-performance vector database for semantic search
• ✅ Auto-ingestion: Automatic document processing for seamless onboarding
• ✅ Robust Error Handling: Fallback mechanisms for reliable operation
• ✅ Debug Logging: Comprehensive logging for troubleshooting and transparency
• ✅ Health Monitoring: Detailed system status for operational excellence

User Experience Improvements

• ✅ Dynamic Height: Responsive containers preventing content truncation
• ✅ Agent-Specific UI: Tailored interfaces for each professional development stage
• ✅ Loading Indicators: Contextual loading messages with progress feedback
• ✅ Better Typography: Improved readability for professional content
• ✅ Error Handling: User-friendly error messages and recovery mechanisms

AI Agent Enhancements

• ✅ Welcome Agent: Enhanced onboarding FAQ responses with policy integration
• ✅ Skill Navigator: Improved 30-60 day learning plan generation
• ✅ Progress Companion: Enhanced progress tracking with motivational feedback
• ✅ Streaming: Real-time response streaming for interactive professional development

🚨 Troubleshooting

Common Issues

1. Vector Store Empty: Check AUTO_INGEST=1 and document files
2. Agent Responses: Verify OpenAI API key and Qdrant connection
3. Frontend Truncation: Ensure dynamic height containers are working
4. Streaming Issues: Check network connectivity and proxy configuration

Debug Commands

# Check vector store status
curl http://localhost:8000/healthz

# Test agent responses
curl -X POST http://localhost:8000/agents/onboarding/stream \
  -H "Content-Type: application/json" \
  -d '{"msg": "What is the leave policy?"}'

# Check document count
python -c "from rag import get_document_count; print(get_document_count())"

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Aurora - AI ecosystem empowering young professionals from onboarding to leadership with personalized guidance, skills growth planning, and transparent feedback